Abstract
The purpose of this study is to extend the capabilities of multispectral optical imaging techniques for morphological characterization of irregular particles, encountered in several applications. We have utilized the Gaussian random ellipsoid model to quantize the shape elongation and deviation in complex-shaped particles. Compared with the Gaussian random sphere model, it is more realistic and is applicable to wider range of irregular particles with minimum possible of free parameters. By processing the reduced light scattering spectra, the proposed inverse technique can simultaneously recover the size, volume fraction, elongation, and shape deformation of particles that are randomly oriented within a colloidal suspension. The discrete dipole approximation is employed in forward light scattering calculations and input synthetic data generation. To investigate the robustness of our algorithm in coping with real scenarios, we have added different levels of noise to the observed scattering spectra. The results demonstrate the potential of our technique to non-invasively recover the morphology parameters of irregular particles.